gamiﬁcation of collaborative learning scenarios: an

Gamification of Collaborative Learning Scenarios: AnOntological Engineering Approach to Deal with Motivational

Problems in Scripted Collaborative LearningGeiser C. Challco1, Seiji Isotani1

1ICMC - University of Sao Paulo (USP), Sao Carlos – SP, Brazil

[email protected], [email protected]

Abstract. To date, one of the major problems in the field of Computer-SupportedCollaborative Learning (CSCL) is to engage students in meaningful interac-tions. Past research has addressed this issue by using scripts, a procedure thathumans and computers can use to reduce extraneous interactions. Nevertheless,it has caused negative learning outcomes and motivational problems over time.Thus, in our work, we propose an innovative approach that combines gamifi-cation, a strategy to increase engagement, and personalization to improve thechances of successful learning. Both gamification and personalization of col-laborative learning (CL) are complex tasks that require a specific organizationof knowledge about game-design and theories of human learning. To deal withthis, we formalize such knowledge into an ontology and, on top of it, we builtcomputational tools, procedures, and mechanisms to support well-thought-outgamified CL activities. We conducted three full-scale empirical studies during asemester in real situations with around 60 students each. Findings of these stud-ies have shown statistically significant differences between our approach and thecurrent state of the art. We found positive effects of our approach on students’intrinsic motivation and skill/knowledge acquisition with a strong positive cor-relation among them. To the best of our knowledge, this is the first work thatdeveloped an effective and efficient method to deal with motivational problems.

1. IntroductionIn CSCL, scripts orchestrate and structure the CL promoting meaningful, fruitful and sig-nificant interactions among students [Fischer et al. 2013]. Despite these benefits, motiva-tional problems may occur in scripted CL. The lack of motivation, also known as amo-tivation, may be caused by the sense of obligation to following an unwilling sequenceof interactions, the lack of interest in content-domain, and the learners’ preference towork individually. The demotivation problem, as the loss of initial motivation, may becaused by a coercion degree of scripts when there is a lack of choice over their sequenceof interactions [Dillenbourg 2002], difficulty to perform structured tasks [Isotani 2009],and when they are executed for a long time [Schmitt and Weinberger 2018]. Motiva-tional problems degrade the level of participation [Wu et al. 2014], persistence, and effort[Weinberger et al. 2005], which may cause negative learning outcomes, such as superfi-cial interactions, and a low level of knowledge elaboration [Xie and Ke 2009].

To deal with motivational problems, Gamification has been pointed out by manyresearchers and practitioners as a promising approach to engage students in educationalcontexts [Koivisto and Hamari 2018]. However, gamification is too context-dependent

DOI: 10.5753/cbie.wcbie.2019.981 981

Anais dos Workshops do VIII Congresso Brasileiro de Informática na Educação (WCBIE 2019)VIII Congresso Brasileiro de Informática na Educação (CBIE 2019)

[Richards et al. 2014]. Thus, the gamification is a non-trivial task, and when the gameelements are not properly applied by tailoring them for each situation, gamification maycause negative effects [Toda et al. 2018], such as detrimental on students’ interest, cheat-ing, embarrassment, and a lack of credibility.

A personalization of gamification for CL scenarios needs to rely on knowledgefrom game design, and from theories and practices of motivation and human behaviorbecause motivation varies from individual to individual, from situation to situation, and itvaries in amounts and types [Deci and Ryan 2010]. Such knowledge hereinafter referredto as theories and practices of gamification, currently lack formal vocabulary and com-mon definitions. This fact hinders the creation of models/frameworks that can be appliedby computational systems to help instructional designers in the gamification of CL ses-sions based on these theories. In this sense, developing a systematic way to represent in anunambiguous way the knowledge from theories and practices of gamification is necessaryto obtain these computational systems. Furthermore, this knowledge must be properlyaligned with knowledge from instructional design and theories of human learning to en-sure the accomplishment of pedagogical goals through CL scenarios.

Ontologies have been consolidated as the most advanced technology to support theexplicit representation of knowledge in a common understandable and shareable mannerfor computers and humans [Mizoguchi and Bourdeau 2016]. Through this representation,ontologies can be used to delineate concepts from theories and practices of gamificationwithout ambiguities, and by taking advantages of the Internet interconnection, ontologieswith these concepts can be used as a language to share their understandings and interpre-tations. Thereby, our main research objective was to develop a solution that integratesgamification and ontologies to deal with motivational problems in scripted CL. The nextsections are structured as follows: firstly, Sec. 2 presents the related works. Sec. 3 presentsour proposed solution to deal with motivational problems in scripted CL. Sec. 4 describesthe research methodology. Sec. 5 presents the findings obtained in empirical studies tovalidate our approach. Sec. 6 shows the conclusions and contributions of our work.

2. Related WorksDealing with motivational problems in scripted CL is a challenge addressed by the CSCLcommunity for many years [Dillenbourg 2002]. One traditional approach is the use ofinstructional design models that focus on motivation. These models, such as the ARCSmodel [Keller 2009], the time continuum model [Wlodkowski and Ginsberg 2017], andthe taxonomy of intrinsic motivations for learning [Malone and Lepper 1987], can be usedas guidelines to define a better way in which learning environments, activities, content,and resources should be developed to motivate the students to follow an effective and ef-ficient CL process. Another traditional approach to deal with motivational problems inscripted CL is the the regulation of affective states through motivational dialogs. Thesemotivational dialogs are messages, such as “Let’s keep going”, and computational sys-tems known as Affective feedback systems can be developed to engender these dialogsbased on the identification of learners’ affective states. Thus, for example, the intelligentsystem “Guru” [Olney et al. 2012] was developed to deal with the demotivation duringcollaborative lectures. [Tian et al. 2014] built a system that sends advice in peer-learningactivities to regulate the participants’ emotions when it identifies boredom, frustration orfury. Finally, as the last example, we have emotcontrol [Feidakis et al. 2014], a tool in

982


which the participants indicate their current affective states, and based on these states, thetool engenders motivational dialogs in form of empathic dialogs during the CL process.

Instructional design models and affective feedback systems are considered tra-ditional approaches because they are based on the assumption that the good quality ofinstructional materials and the content-domain by itself is compelling and interesting foreveryone. This assumption ignores the fact that some learners may dislike the content-domain, environments, activities or resources used in the CL process. In this sense, effortsof the CSCL community has been directed to finding new innovative solutions that, be-sides to motivate and engage students during the entire CL process, are not completely tiedto the domain-content and desired to learn to work in groups. In this direction, gamifica-tion has picked the attention of many several researchers. However, in the literature, onlyone approach was found in which, to personalize the gamification of CL scenarios, ma-chine learning is used to identify individual profiles of gamification [Knutas et al. 2017].However, this solution is not oriented to deal with motivational problems in the scriptedCL, and its purpose is to increase the communication of the participants.

3. An Ontological Engineering Approach to Gamify CL ScenariosFig. 1 shows our ontological approach to gamify CL scenarios proposed as a solution todeal with motivational problems in scripted CL.

Fig. 1. Ontological engineering approach to gamify CL scenarios.

This approach consists of three major stages described as follows: The stage (1)is the application of ontology engineering in the theories and practices of gamificationto represent the knowledge about how to gamify CL scenarios into an ontology OntoGa-CLeS1 - part of this ontology is shown in Fig. 2 in which there are: (A1) the ontologicalstructure to represent the Yee Socializer role (player role), and (A2) the ontological struc-ture to represent the Gamified Cognitive Apprenticeship Sessions for Yee Achiever/YeeSocializer (ont-gamified CL session). The stage (2) is the development of computationalmechanisms and procedures based on our ontology to support the gamification of CLscenarios. In the stage (3), this computational support aims to be given through tools inform of advice and recommendations to obtain tailored gamified CL sessions known asontology-based gamified CL sessions (called in short, ont-gamified CL sessions).

Fig. 2 shows the computational mechanisms and procedures developed on top ofour ontology, and they are: a conceptual flow to gamify CL sessions shown in Fig. 2 (B),a pseudo-algorithm to set player roles illustrated in Fig. 2 (1), a procedure to design theCL gameplay illustrated in Fig. 2, and a reference architecture of intelligent theory-awaresystems to gamify CL sessions shown in Fig. 2 (C). As part of the procedure to design the

1https://github.com/geiser/ontogacles

983


https://github.com/geiser/ontogacles

CL gameplay, to define the values of rewards for each interaction of the scripted CL, wealso developed a GIMF model - illustrated in Fig. 2 (2b) - that integrates the balance ofchallenge/ability proposed by the flow theory with the Learner’s Growth Model (LGM)proposed by [Isotani 2009]. We also proposed a theory-aware tool that, in a prescriptiveway, uses the WAY-knowledge base to represent the design rationale to gamify a non-game event, such as the Setting up learning context type CA illustrated in Fig. 2 (2c).

Fig. 2. Mechanisms/procedures developed on top of the ontology OntoGaCLeS.

4. Research Methodology

As our work aims to achieve theoretical and empirical contributions, the mixed researchmethodology proposed by [Glass et al. 2002] was carried out in four iterative phases: in-formational, propositional, analytical and evaluation. Through scientific (observing theworld) and engineering (observing existing solutions) research methods, in the informa-tional phase, we achieved the theoretical contributions referred as the identification of rel-evant concepts from theories and practices of gamification for dealing with motivationalproblems in scripted CL. During the propositional phase, through ontology engineering,we formalized ontological structures into the ontology OntoGaCLeS to represent the rel-evant concepts identified in the informational phase. On top of this ontology, during theanalytical and evaluation phases, we achieved the empirical contributions referred to asthe development of computational tools, methods, and procedures to support the well-though-out gamification of CL sessions. In this phase, we also validated our approach

984


through empirical studies conducted in real situations in which we assessed the effective-ness and efficiency of our approach to deal with motivational problems in scripted CL.

5. Empirical Studies and ResultsOne-pilot and three full-scale empirical studies were carried out to: analyze “the ef-fects of ont-gamified CL sessions on the students’ motivation and learning outcomes”for “validating our ontological engineering approach” concerning “the effectiveness andefficiency to deal with motivational problems in scripted CL.” The effectiveness was eval-uated in the pilot, 1st and 2nd studies by comparing ont-gamified CL sessions (experi-mental condition) against non-gamified CL sessions (control condition) using as metricsstudents’ motivation and learning outcomes. Efficiency was evaluated in the 3rd studyby comparing students’ motivation and learning outcomes in ont-gamified CL sessions(experimental condition) against gamified CL sessions that use the current state of the artapproach to gamify an activity (control condition, in short, called w/o-gamified CL ses-sions). These studies were carried out in real situations with undergraduates students inthe course of Introduction to Computer Science offered at the University of Sao Paulo.Thereby, the CL sessions for the empirical studies have been instantiated using a CSCLscript inspired by the Cognitive Apprentice theory, and to three CL activities with thesubjects of: conditional structures for the 1st empirical study (with N = 62 students);loop structures for the pilot study (with N = 39) and 2nd empirical study (with N = 58);and recursion for the 3rd empirical study (with N = 59).

All the empirical studies were executed with a 2 × 2 factorial design, and em-ploying pre-test, intervention, and post-test phases. During the pre-test, multiple-choiceknowledge questionnaires and programming problem tasks were applied to the studentsto determine their CL roles (master or apprentice) in the CL sessions. In the interventionphase, this CL role assignment was performed through the theory-driven algorithm pro-posed in [Isotani 2009], and with a randomized assignment of students for the type of CLsession (non-gamified CL session - as exemplified in Fig. 3 (a), w/o-gamified CL session -as exemplified in Fig. 3 (b), and ont-gamified CL sessions - as exemplified in Fig. 3 (c), (d)and (e). For the pilot, 1st and 2nd empirical studies, the ont-gamified CL sessions wereinstantiated from the structures “Gamified Cognitive Apprenticeship Scenario for Mas-ter/Yee Achiever and Apprentice/Yee Achiever” and “Gamified Cognitive ApprenticeshipScenario for Master/Yee Socializer and Apprentice/Yee Socializer.” Thus, as exemplifiedin Fig. 3 (c), for an apprentice student who plays the Yee Achiever role because he/she hadmore liking for achievement-components than for social-components, an individual com-petition was set up to him/her by using: individual achievements - shown in Fig. 3 (c1) -with badges based on levels of individual participation - shown in Fig. 3 (c2), individualpoint-systems - shown in Fig. 3 (c3), and a leaderboard of individual rankings - shown inFig. 3 (c4). For an apprentice student who plays the Yee Socializer role because he/shehad more liking for social-components than for achievement-components, as exemplifiedin Fig. 3 (d), a collaborative competition was established using: team achievements -shown in Fig. 3 (d1) - with badges based on levels of collaborative participation - shownin Fig. 3 (d2), point-systems for groups - shown in Fig. 3 (d3), and a leaderboard withrankings for teams - shown in Fig. 3 (d4).

For the 3rd empirical study, the ont-gamified CL sessions were instantiated fromthe structures “Gamified Cognitive Apprenticeship Scenario for Master/Yee Achiever

985


Fig. 3. Non-gamified, w/o-gamified and ont-gamified CL sessions for an appren-tice student in the scripted CL based on the Cognitive Apprentice.

and Apprentice/Yee Achiever” and “Gamified Cognitive Apprenticeship Scenario forMaster/Social Achiever and Apprentice/Social Achiever.” Thus, students who likeachievement-components and social-components were assigned to play the SocialAchiever role in gamified CL sessions that promote a collaborative-competition, as ex-emplified in Fig. 3 (e) for an apprentice student. In this kind of ont-gamified CL ses-sions, the collaborative-competition is supported by individual achievements - shown inFig. 3 (e1), and team achievements - shown in Fig. 3 (e2), with badges based on indi-vidual and collaborative participation - shown in Fig. 3 (e3). Point-systems for groups -shown in Fig. 3 (e4), a leaderboard with individual rankings - shown in Fig. 3 (e5), anda leadearboard with rankings for teams - shown in Fig. 3 (e6) support the collaborative-competition. In this 3rd empirical study, students who only like achievement-componentswere assigned to play the Yee Achiever role - as was detailed above and exemplified inFig. 3 (c) for an apprentice student. W/o-gamified CL sessions - as exemplified in Fig. 3(b) - are close/similar to the ont-gamified CL sessions for Yee Achiever. In both types ofCL sessions, individual achievements with badges based on individual participation - asexemplified in Fig. 3 (b1) and (b2), individual point-system with a leaderboard with indi-vidual ranking support the individual competition - as exemplified in Fig. 3 (b3) and (b4).The only difference between w/o-gamified CL sessions and ont-gamified CL sessions forYee Achiever is the value of points given as rewards for each interaction in the CL pro-cess. In w/o-gamified CL sessions, these rewards were defined by the teacher without

986


using the GIMF model - as shown in Fig. 3 (b3), whereas these rewards were defined bythe GIMF model in ont-gamified CL sessions - as shown in Fig. 3 (c3).

In the post-test phase, multiple-choice knowledge questionnaires and program-ming problem tasks were applied to the students to determine the learning outcomes of CLsessions as gains of skill/knowledge. During this phase, motivation surveys, IMI (IntrinsicMotivation Inventory) and IMMS (Instructional Materials Motivation Survey), were alsoapplied for measuring the students’ motivation as measurement of the intrinsic motiva-tion (interest/enjoyment, perceived choice, pressure/tension, and effort/importance) andthe level of motivation (attention, relevance, and satisfaction). These motivation surveysare well-known and broadly applied in the research community, and the reliability of theseinstruments was validated with a Cronbach’s α = 0.894 (IMI) and α = 0.909 (IMMS).

Through two-way ANOVA tests and Tukey posthoc, we found significant differ-ences in the students’ motivation and their gains in skill/knowledge with relation to thetype of CL session and the CL role played by students. Besides to perform these analyseswith the scores obtained by the students through non-parametric and parametric tests, theestimates latent traits of motivation and gains in skill/knowledge were measured throughIRT-models. Figure 4 only summarizes results of the statistical analysis based on theestimates from IRT-models. These results only correspond to the statistically significantdifference found at the level of < 0.05 (p-adj), and with relevance for the claimed con-tributions described in this paper. Findings of the 1st empirical study were published in[Challco et al. 2019], and all scripts, instruments, data from experiments, and their sta-tistical analysis are freely available at http://bit.ly/2m6L7ko, contributing to theopen-access of knowledge and dissemination and validation of scientific results.

Fig. 4. Summary of statistical significance findings in the empirical studies.

In regards to the effectiveness for dealing with motivational problems, the resultsof the pilot, first and second empirical studies indicate that our approach, which cre-ates ont-gamified CL sessions, have positive impacts on the students’ intrinsic motivationand perceived choice if compared with non-gamified CL sessions. Furthermore, Whenthe content was conditional structures, we had extra benefits related to the metrics pres-sure/tension and effort/importance to complete CL tasks. When the content was loopstructures, in addition to the positive effects discussed previously, our approach has in-creased students’ interest/enjoyment related to working in CL activities. About the effi-

987


http://bit.ly/2m6L7ko

ciency of dealing with motivational problems, results of the third empirical study indicatesthat the students’ intrinsic motivation, perceived choice, effort, and importance were sta-tistically significantly greater in when working in ont-gamified CL sessions (experimentalcondition) if compared with w/o-gamified CL sessions (control condition). Moreover, thegains of skill/knowledge obtained by students in ont-gamified CL session were greaterthan in w/o-gamified CL sessions.

Finally, the statistically significant results of Spearman’s rank-order correlationtests in ont-gamified CL sessions show that the intrinsic motivation and perceived choiceare strongly positively correlated with students’ gains of knowledge/skill. This fact sug-gests that the improvement of learning outcomes in ont-gamified CL sessions was a con-sequence of the positive effects in students’ motivation caused by the adequate personal-ization of game elements of our approach.

6. Conclusions and ContributionsOur main contribution has been to develop and empirically evaluate an approach to ad-dress, effectively and efficiently, the motivational problems that may occur in scriptedCL. This result is highly relevant for the CSCL community that, in the past few years,has made efforts to find solutions to motivate and engage students when they work incollaborative tasks. To the best of our knowledge, this is the first approach that has beenable to empirically demonstrate the benefits of ontologies and personalized gamificationin students’ learning and their motivation during scripted CL.

The practical implication of this contribution is that gamification now can be ad-equately and semi-automatically used as a good alternative to the current state of the artsolutions that aim to motivate and engage students when working in CL activities. In par-ticular, one of the benefits of our approach is that students without interest/desire to learnare positively affected. This benefit is ensured in our ont-gamified CL sessions by theproper alignment of pedagogical objectives with the individual motivational strategies,player roles, and game elements in the ontological structure to represent gamified CLscenarios. Another benefit of our approach is that, through the application of PersuasiveGame Design (PGD), students can be motivated during the CL process to follow interac-tions patterns indicated in CSCL scripts. This benefit is ensured in our ont-gamified CLsessions by the proper connection between the PGD and the design of CL process that isrepresented in descriptive and prescriptive ways in the ontology OntoGaCLeS.

Our work also contributes to building the theoretical foundations of gamifica-tion. By demonstrating that, through our approach, we can obtain better intrinsic mo-tivation than with current approaches, we provided strong empirical evidence that sup-ports the need of using a tailored-based gamification approach instead of a one-size-fits-all approach. Furthermore, the positive effects of our approach on the students’ gainsof skill/knowledge with a strong positive correlation with the intrinsic motivation sug-gests that, to achieve the full potential and benefits of gamification, it is highly importantto focus on the use of well-grounded theoretical knowledge extracted from theories andpractices of gamification rather than only focus on building adaptive mechanisms.

We also offer a significant contribution to bridging the gap between two differentcommunities, namely CSCL and gamification, that have been working to deal with mo-tivational problems during the learning process. To do that, our formalization of knowl-

988


edge that resulted in the ontology OntoGaCLeS has been essential. It brings practicaland methodological implications. By employing the ontological structures (detailed inChap. 3 and Chap. 4 of the thesis [Challco 2018]), we can represent the knowledge fromgame design and theories of motivation and human learning to build gamification modelsthat support the personalization of game elements (e.g. personalization of game elementsbased on player type models). Based on our ontology, we built computational proceduresand mechanisms to support the semi-automatic design of well-thought-out gamified CLscenarios. These procedures/mechanisms (detailed in Chap. 5 and Chap. 6 of the thesis[Challco 2018]) constitute an important contribution that aims to facilitate the develop-ment of a new generation of intelligent-theory aware systems to gamify CL sessions.

From a methodological perspective, our research work, briefly summarized in thispaper, demonstrated that, through ontology engineering, we can identify and extract thenecessary knowledge to properly gamify CL scenarios for dealing with motivational prob-lems. Thus, employing the same research methodology of the thesis, we can extendthe ontological structures of the ontology OntoGaCLeS to deal with other motivationalproblems that may also occur in other CL situations, such as peer-assessment, informalCL groups, and PBL (Problem-Based Learning). By reusing the ontological structuresproposed in OntoGaCLeS, and employing our research methodology, new ontologicalstructures can be built to solve other challenging motivational problems related to otherresearch areas, such as individual learning, blended learning, project management, gover-nance, and entrepreneurship.

ReferencesChallco, G. C., , Bittencourt, I. I., and Isotani, S. (2019). The effects of ontology-based

gamification in scripted collaborative learning. In 19th International Conference onAdvanced Learning Technologies (ICALT-IEEE), pages 123–128. IEEE.

Challco, G. C. (2018). Gamification of collaborative learning scenarios: an ontologi-cal engineering approach to deal with the motivation problem caused by computer-supported collaborative learning scripts. Ph.D. dissertation, USP.

Deci, E. L. and Ryan, R. M. (2010). Self-Determination. The Corsini Ency. of Psychol-ogy. Wiley & Sons, Inc.

Dillenbourg, P. (2002). Over-scripting CSCL: The risks of blending collaborative learningwith instructional design. In Three worlds of CSCL. Can we support CSCL?, pages 61–91, Nederland. Open Universiteit.

Feidakis, M., Daradoumis, T., Caballe, S., and Conesa, J. (2014). Embedding emotionawareness into e-learning environments. International Journal of Emerging Technolo-gies in Learning (iJET), 9(7):39–46.

Fischer, F., Kollar, I., Stegmann, K., Wecker, C., and Zottmann, J. (2013). Collaborationscripts in computer-supported collaborative learning. The international handbook ofcollaborative learning, pages 403–419.

Glass, R., Vessey, I., and Ramesh, V. (2002). Research in software engineering: an anal-ysis of the literature. Information and Software Technology, 44(8):491–506.

Isotani, S. (2009). An Ontological Engineering Approach to Computer-Supported Col-laborative Learning. PhD thesis, Osaka University, Japan.

989


Keller, J. M. (2009). Motivational design for learning and performance: The ARCS modelapproach. Springer Science & Business Media.

Knutas, A., van Roy, R., Hynninen, T., Granato, M., Kasurinen, J., and Ikonen, J. (2017).Profile-Based Algorithm for Personalized Gamification in Computer-Supported Col-laborative Learning Environments. In Computer-Supported Collaborative LearningEnvironments. CEUR Workshop Proceedings.

Koivisto, J. and Hamari, J. (2018). The rise of motivational information systems: A reviewof gamification research. Int. Journal of Information Management, 45:191–210.

Malone, T. W. and Lepper, M. R. (1987). Making learning fun: A taxonomy of intrinsicmotivations for learning. volume 3, pages 223–253. Erlbaum, Hillsdale, N.J.

Mizoguchi, R. and Bourdeau, J. (2016). Using Ontological Engineering to OvercomeAI-ED Problems: Contribution, Impact and Perspectives. Int. Journal of ArtificialIntelligence in Education, 26(1):91–106.

Olney, A. M., D’Mello, S., Person, N., Cade, W., Hays, P., Williams, C., Lehman, B., andGraesser, A. (2012). Guru: A Computer Tutor That Models Expert Human Tutors. InIntelligent Tutoring Systems, LNCS, pages 256–261. Springer.

Richards, C., Thompson, C. W., and Graham, N. (2014). Beyond Designing for Moti-vation: The Importance of Context in Gamification. In Proceedings of the First ACMSIGCHI Annual Symposium on Computer-human Interaction in Play, CHI PLAY ’14,pages 217–226, Canada. ACM.

Schmitt, L. J. and Weinberger, A. (2018). Computer-Supported Collaborative Learning:Mediated and Co-Present Forms of Learning Together. In Second Handbook of Infor-mation Technology in Primary and Secondary Education, pages 1–15. Springer, Cham.

Tian, F., Gao, P., Li, L., Zhang, W., Liang, H., Qian, Y., and Zhao, R. (2014). Recognizingand regulating e-learners’ emotions based on interactive Chinese texts in e-learningsystems. Knowledge-Based Systems, 55:148–164.

Toda, A., Valle, P., and Isotani, S. (2018). The Dark Side of Gamification: An Overview ofNegative Effects of Gamification in Education, volume 832 of 1st International Work-shop on Social, Semantic, Adaptive and Gamification Techniques and Technologies forDistance Learning, HEFA 2017. Springer Verlag.

Weinberger, A., Ertl, B., Fischer, F., and Mandl, H. (2005). Epistemic and social scriptsin computer-supported collaborative learning. Instructional Science, 33(1):1–30.

Wlodkowski, R. J. and Ginsberg, M. B. (2017). Enhancing Adult Motivation to Learn: AComprehensive Guide for Teaching All Adults. Wiley.

Wu, K., Vassileva, J., Sun, X., and Fang, J. (2014). Motivating Wiki-Based CollaborativeLearning by Increasing Awareness of Task Conflict: A Design Science Approach. InCollaboration and Technology, pages 365–380. Springer.

Xie, K. and Ke, F. (2009). How Does Students’ Motivation Relate to Peer-moderatedOnline Interactions? In Proceedings of the 9th International Conference on ComputerSupported Collaborative Learning - Volume 1, CSCL’09, pages 242–251, Rhodes,Greece. International Society of the Learning Sciences.

990


gamiﬁcation of collaborative learning scenarios: an

Documents